The Harvey-Bradshaw Index Adapted to a Mobile Application Compared with In-Clinic Assessment: The MediCrohn Study

Abstract

Objectives: Mobile apps are useful tools in e-health and self-management strategies in disease monitoring. We evaluated the Harvey-Bradshaw index (HBI) mobile app self-administered by the patient to see if its results agreed with HBI in-clinic assessed by a physician. Methods: Patients were enrolled in a 4-month prospective study with clinical assessments at months 1 and 4. Patients completed mobile app HBI and within 48 h, HBI was performed by a physician (gold standard). HBI scores characterized Crohn's disease (CD) as remission <5 or active >/=5. We determined agreement per item and total HBI score and intraclass correlation coefficients (ICCs). Bland-Altman plot was performed. HBI changes in disease activity from month 1 to month 4 were determined. Results: A total of 219 patients were enrolled. All scheduled assessments (385 pairs of the HBI questionnaire) showed a high percentage of agreement for remission/activity (92.4%, kappa = 0.796), positive predictive value (PPV) for remission of 98.2%, and negative predictive value of 76.7%. High agreement was also found at month 1 (93.15%, kappa = 0.82) and month 4 (91.5%, kappa = 0.75). Bland-Altman plot was more uniform when the HBI mean values were <5 (remission). ICC values were 0.82, 0.897, and 0.879 in all scheduled assessments, 1 and 4 months, respectively. Conclusions: We found a high percentage of agreement between patients' self-administered mobile app HBI and in-clinic physician assessment to detect CD activity with a remarkably high PPV for remission. The mobile app HBI might allow a strict control of inflammation by remote monitoring and flexible follow-up of CD patients. Reduction of sanitary costs could be possible.